compooter-enhensa/common.h

#include <stdint.h>   /* uint32_t etc */
#include <stdlib.h>   /* exit */
#include <stdio.h>    /* perror, fprintf */
#include <fcntl.h>    /* open */
#include <unistd.h>   /* read */
#include <sys/stat.h> /* fstat */

typedef uint8_t  u8;
typedef uint16_t u16;
typedef uint32_t u32;
typedef uint64_t u64;

typedef int8_t  s8;
typedef int16_t s16;
typedef int32_t s32;
typedef int64_t s64;

#define MAX_INSTRUCTIONS 1024
#define MAX_LABELS 128

enum regs_w {
    AX = 0,
    CX = 1,
    DX = 2,
    BX = 3,
    SP = 4,
    BP = 5,
    SI = 6,
    DI = 7,
};

enum regs_h {
    AL = 0,
    CL = 1,
    DL = 2,
    BL = 3,
    AH = 4,
    CH = 5,
    DH = 6,
    BH = 7,
};

enum mov_mod {
    MOD_MEMORY    = 0x00,
    MOD_MEMORY_8  = 0x01,
    MOD_MEMORY_16 = 0x02,
    MOD_REGISTER  = 0x03,
};

enum mov_opcode {
    OPCODE_REG_RM  = 0x88,
    OPCODE_IMM_RM  = 0xC6,
    OPCODE_IMM_REG = 0xB0,
    OPCODE_MEM_ACC = 0xA0,
    OPCODE_ACC_MEM = 0xA2,
};

enum mov_opcode_mask {
    MASK_REG_RM  = 0xFC,
    MASK_IMM_RM  = 0xFE,
    MASK_IMM_REG = 0xF0,
    MASK_MEM_ACC = 0xFE,
    MASK_ACC_MEM = 0xFE,
    
    MASK_ADD_SUB_CMP = 0xFC,
};

struct rm {
    s8  reg1; // -1 here means it's an immediate
    s8  reg2; // -1 means no second
    s16 disp; // 0 means no displacement
    s16 imm;  // if it's not an immediate - imm indicates if there is a displacement
};

struct instruction {
    u8 size;
    u8 opcode;
    u8 wide;
    u8 swap;
    
    /* for mov, add, sub, cmp */
    struct rm dst;
    struct rm src;
    
    /* for control-flow instructions */
    s8 jump_offset;
};

static const s8 NO_REGISTER = -1;
static const u8 ANTIOP_MASK = 0xC4;

static const char *REGS_W0[] = { "al", "cl", "dl", "bl", "ah", "ch", "dh", "bh" };
static const char *REGS_W1[] = { "ax", "cx", "dx", "bx", "sp", "bp", "si", "di" };

static const s8 MOD_PAIRS_FIRST[]  = { BX, BX, BP, BP, SI, DI, BP, BX };
static const s8 MOD_PAIRS_SECOND[] = { SI, DI, SI, DI, NO_REGISTER, NO_REGISTER, NO_REGISTER, NO_REGISTER };

static const char *OP_NAMES[] = {
    [0x0] = "add",
    [0x5] = "sub",
    [0x7] = "cmp",
    [0xF] = "mov", /* Source? I made it up */
    
    [0x74] = "je",
    [0x7C] = "jl",
    [0x7E] = "jle",
    [0x72] = "jb",
    [0x76] = "jbe",
    [0x7A] = "jp",
    [0x70] = "jo",
    [0x78] = "js",
    [0x75] = "jne",
    [0x7D] = "jnl",
    [0x7F] = "jg",
    [0x73] = "jnb",
    [0x77] = "ja",
    [0x7B] = "jnp",
    [0x71] = "jno",
    [0x79] = "jns",
    
    [0xE2] = "loop",
    [0xE1] = "loopz",
    [0xE0] = "loopnz",
    [0xE3] = "jcxz",
};

#define CRASH_NOTIMPL() do { fprintf(stderr, "line: %d. not implemented\n", __LINE__); exit(1); } while (0) 
#define CRASH_SHOULDNOT() do { fprintf(stderr, "line: %d. should not happen\n", __LINE__); exit(1); } while (0) 

static u64
file_size(int fd)
{
    struct stat st = { 0 };
    
    if (fstat(fd, &st) == -1) {
        perror("fstat");
        exit(1);
    }
    
    return(st.st_size);
}

static u8 *
read_file(int fd, u64 size)
{
    u8 *buf = malloc(size);
    
    if (!buf) {
        perror("malloc");
        exit(1);
    }
    
    int rt = read(fd, buf, size); 
    
    if (rt == -1) {
        perror("read");
        exit(1);
    }
    
    if (rt != (int) size) {
        fprintf(stderr, "short read: %d < %lu\n", rt, size);
        exit(1);
    }
    
    return(buf);
}